Their AI pipelines consist of a sequence of interrelated processing and optimization steps, allowing for end-to-end optimization of their customer’s heat networks. This results in a significant reduction of heat losses and peak optimization, and (predictive) maintenance. An additional benefit is a significant reduction of CO2 emissions.
Gradyent selected UbiOps for deploying, running and managing their AI algorithms. Together we deployed and configured an instance of UbiOps OnPremise in Gradyent’s own Amazon Web Services (AWS) environment after which Gradyent Data Scientists could start deploying their AI pipelines and models within UbiOps. For this Gradyent used the UbiOps WebApp as well as the UbiOps API and Client Libraries, which allowed for scripted pipeline and model deployment. The process of deploying and managing Gradyent’s AI pipelines was significantly shorter and DevOps costs were reduced to a minimum.
Using UbiOps, our team was able to orchestrate all processing steps and data flows seamlessly so we could start using our models while minimizing DevOps costs.
Ivo Fugers
AI Lead at Gradyent
As is common in Data Science, Gradyent is constantly improving their models and algorithms. UbiOps allows them to upload and replace models within existing pipelines in a matter of minutes.
Using the auto-scaling feature of UbiOps, server compute costs are optimized and reduced to a minimum, while models can depend on getting all the server power they need to complete their tasks in the desired time.
Driven by the demand of the growing number of Gradyent customers, significant improvements have been made to reduce installation time to a minimum. New releases of UbiOps are quickly deployed on the Gradyent clusters using installation scripts. This reduces downtime to a minimum and ensures that Gradyents models continue running as expected.
With UbiOps we are able to fasten the time to market of our models significantly. Building such an infrastructure ourselves in our own Cloud would cost us at least half a year, while we want to quickly bring value to our clients.
The extensive API documentation gives us the ability to build pipelines from scratch, without having to go through a 3 week training program.
Hervé Huisman
Founder & CEO at Gradyent
Gradyent makes use of UbiOps OnPremise; whereby UbiOps is deployed in their own AWS environment. Providing the necessary data processing security.
UbiOps handles over 15.000 model calls every day.
Using the UbiOps API and Client Libraries Gradyent was able to automate and script many of the tasks needed to train, operationalize, monitor, version and maintain their data science pipelines.
Where deployment of models previously took months of time and effort, with UbiOps this process was reduced to hours.
© 2023 UbiOps.
